Simulation of Regional Climate Using a Limited Area Model Nested in a General Circulation Model

A Limited Area Model (LAM) is nested in a General Circulation Model (GCM) to simulate the January climate over the western United States. In the nesting procedure, the GCM output is used to provide the initial and lateral atmospheric boundary conditions necessary to drive the LAM. In this approach, the GCM is used to simulate realistic large-scale atmospheric behavior over an area of interest and the LAM to describe the effect of local, sub-GCM grid scale forcings (such as those induced by the complex western United States topography) on regional patterns of climatic variables. Two versions of the National Center for Atmospheric Research (NCAR) Community Climate Model the seasonal CCM1 at 4.5° × 7.5° (R15) and 2.89° × 2.89° (T42) latitude-longitude resolution are used to drive a version of the Pennsylvania State University/NCAR mesoscale model (MM4 at 60 km resolution), which includes sophisticated soil hydrology calculations. The CCM1 large-scale January climatology over the region is analyzed first. Comparison with large-scale observations shows that geopotential height, zonal wind, temperature, relative humidity, cloudiness, precipitation and storm frequencies over the western United States and adjacent oceans are realistically simulated by both the T42 and R15 models. The T42 model, however, reproduces storm frequencies and strength and position of the jet stream better than the R15 model. A number of month-long January simulations were performed using both the R15 and T42 model outputs to drive the MM4. The large-scale average circulations over the western United States simulated by the nested MM4 are not substantially different from those of the driving CCM1, both when outputs from the R15 and T42 versions are used to drive the MM4. Owing to the more realistic topography in the MM4, the nested model system produces better regional detail of precipitation and temperature distribution than the CCM1 alone. Temperature and precipitation means, as well as frequencies of daily precipitation intensifies simulated by the nested MM4, compare well with high resolution observations, particularly in their spatial distribution. Also discussed am results of regional snow cover, cloudiness, and soil hydrology calculations included in the MM4.